Human tumors both primary and metastatic continue to proliferate in spite of the presence of T and B lymphocytes, plasma cells, natural killer cells and macrophages within and around the tumor. Functional studies of the tumor infiltrating leukocytes (TIL) demonstrate that these cells have the potential to respond specifically to tumor cells and to kill them, but that this anti-tumor activity is often subverted by one of several possible mechanisms that leads to defects in the recognition, activation or effector phases of TIL. The goal of this grant is to overcome these defects in human TIL by modifying the tumor or local tumor environment such that it results in the activation and mobilization of TIL to kill human tumor cells. Several approaches have been taken with animal tumor models to augment anti-tumor immunity including the engineering of tumor cells with cytokine genes, and the introduction of a co-stimulatory molecule into tumors. The results of these animal studies are extremely encouraging and have led to a number of clinical trials. However, this growing enthusiasm for the applicability of these immunotherapeutic protocols to humans should be tempered since there are still very large gaps in our basic knowledge of how these approaches are working in the mouse, and of whether or not they will work in spontaneously arising human tumors. The first aim of this grant is to evaluate IL-2, IL- 12, IL- 15 and GMCSF and combinations of these cytokines for their ability to provoke an optimal anti-tumor response of human TIL in situ. The ability to coengraft scid mice with TIL and tumor by the introduction of non- disrupted pieces of tumor biopsy tissue has made it possible for the first time to monitor the effects of cytokines on TIL activity in situ and to assess their effects upon tumor growth for up to 22 weeks. The second aim is to design and test protocols for the delivery of cytokine genes to tumor cells in vivo such that transient expression of cytokine genes is achieved in a portion of the tumor cells. This is accomplished either by the direct injection of DNA complexed to cationic liposomes into the tumor xenograft in scid mice, or by the intravenous injection of tumor bearing mice with immunospecific DNA-liposome complexes. Aim 3 is to determine whether the in vivo cytokine gene delivery provokes a TIL mediated anti- tumor response. This is achieved by applying the gene delivery protocols to scid mice bearing autologous tumor/TIL xenografts, and to scid mice coengrafted with human peripheral blood leukocytes and allogeneic human lung tumors. The final approach to modifying the human tumor environment focuses upon costimulatory molecules that are expressed on human B and T cells. In this last aim TIL in human tumor xenografts are triggered with antibodies or soluble fusion proteins that react specifically with either CD4O or with CD4O ligand and the effect of this activation upon tumor progression is monitored.